Where the money actually goes in a clinical trial

The standard number for developing a drug is somewhere north of a billion dollars, and the standard explanation is that discovery is hard. That explanation survives mostly because the people repeating it have never sat in a site activation meeting.

Phase II$60M · 2.9y · 29% advance
The valley. Efficacy meets reality and roughly seven in ten programs stop here. Enrollment speed dominates cost.

FIG 1 Cost and duration by phase. Bar width is time, height is spend, so area approximates total cost. Hover or tap a phase. Static version.

Look at where the area sits. Phase III dominates, and almost none of that money is scientific. It goes to sites, monitors, coordinators, and the long tail of finding people who meet the inclusion criteria and are willing to be randomized.

The enrollment constraint

A trial that needs eight hundred patients across ninety sites is not gated by biology. It is gated by how many of those sites can activate, and how quickly each one can find eligible patients in its own catchment.

This is the part of the argument where China becomes unavoidable — but that is the next essay.